Remember the exceptions
The issue has just come up again in reader questions about DNA testing for the direct female line and in Facebook threads… so The Legal Genealogist will once again sound the caution:
Beware the myth of the mtDNA GD0.
That’s a zero at the end, and the GD stands for genetic distance — by definition, “the term used to describe the number of differences or mutations between two sets of … DNA test results.”1
So an mtDNA GD0 means two identical sets of test results: the two people who’ve tested have identical mitochondrial DNA (mtDNA) — the kind of DNA we all have, that we inherit from our mother, that our mother inherited from our grandmother, and our grandmother from our great grandmother all the way back through time in an unbroken female line.2
That identical result is what we hope for when we test the entire mtDNA in what’s called a full mitochondrial sequence. We want that result because, when we get that GD0 match, we can be reasonably sure that — somewhere back in time — we and the person we match will share a common female ancestor.
So… what if we don’t get that GD0 match?
What about the people who match us at a genetic distance of 1 or even 2?
So many times you’ll hear someone say to ignore anything but a GD0 match. Mitochondrial DNA is so stable — it changes so little over the generations — that, we are told, anything that isn’t a perfect match is going to be so far back in time that we’d never get any genealogically valuable data by pursuing it.
Um… not so fast.
That’s the myth of the GD0.
And, like so much else that we encounter in genealogy — and especially in genetic genealogy — it ain’t necessarily so.
Take a look at this chart, of my own family:
The four women identified in yellow have all done full mtDNA testing. All of us descend not just from a common female ancestor, but from a very recent common female ancestor: my grandmother, Opal (Robertson) Cottrell (1898-1995). And since mtDNA is passed from mother to daughter to daughter, all four of us should share identical mtDNA: my grandmother passed her mtDNA to her daughters, my mother and my aunt, my mother’s sister; my mother passed it to my sister and to me; my aunt passed it to my cousin.
We should all be GD0 from each other, right?
But we’re not.
I am a GD1 — a genetic distance of 1 — from my sister. Despite the fact that, yes, we are full blood biological sisters, we do not have perfectly matching mtDNA.
I have something called a heteroplasmy in one part of my mtDNA. That, by definition, is “the presence of a mixture of more than one type of … mitochondrial DNA … within a cell or individual.”3 Where the test expects to see one marker, I show two. That is counted as a genetic difference, and I have one such difference from my sister, who doesn’t have that heteroplasmy.
I’m also a GD1 from my aunt. She doesn’t have that heteroplasmy either. So she and my sister are GD0, and I’m GD1 to each of them.
My cousin, meanwhile, also has a heteroplasmy, but a different one from the one I have. That means she is a GD1 from her own mother and a GD1 from my sister, her first cousin.
And — because each heteroplasmy counts as a genetic difference, and I have one and my cousin has another — my full maternal first cousin and I are a GD2 from each other.
Think about that for a minute.
Would anybody seriously argue that there’s nothing genealogically relevant that two sisters might be able to share with each other — even though those sisters are not GD0 mtDNA matches? Or a mother and daughter who are not GD0? Or first cousins who are GD2 from each other?
So don’t automatically write off those mtDNA matches who aren’t GD0.
Now don’t get me wrong here. I’m not saying to spend a lot of time chasing distant mtDNA matches. They certainly won’t be the ones you focus on first. They may in fact be so far distant that you’ll never find any genealogically relevant information by contacting them.
But then again they may end up being the exceptions… the ones who really are close enough to be not just genealogically relevant but genealogical goldmines… the one who, if you ignore them, could leave you — and them — as victims of the myth of the GD0.
SOURCES
- ISOGG Wiki (http://www.isogg.org/wiki), “Genetic distance,” rev. 31 Jan 2017. ↩
- See ibid., “Mitochondrial DNA tests,” rev. 1 Aug 2017. ↩
- Ibid., “Heteroplasmy,” rev. 31 Jan 2017. ↩
Judy, I wholeheartedly agree. I am a GD2 from the result I can triangulate back to my maternal grandmother and up the matrilineal line. Two 1st cousins and a third cousin who share this line all agree as to what that mtDNA signature should be. I’m the mutant. Since I’m an only child and my mother has been dead for almost two decades, I don’t think I will ever be able to determine if these two mutations occurred between my grandmother and my mother or between my mother and me or if one occurred in each generation. If we encouraged our mtDNA databases to grow we just might find genealogically relevant information.
I have about a dozen GD0 matches. Several are clearly New England lines but without extensive paper trees. Two of us have traced on paper back to different daughters of the same woman, a 1635 immigrant from Devonshire – the mtDNA match supports the accuracy of our paper trees. And one match is a person still in Devonshire. More interesting, perhaps, is that at least one GD1 match has a tentative tree to still a third daughter of the 1635 immigrant. The point here is that mutations do occur and they therefore they have to occur somewhere and sometime between a mother and daughter. There is no reason the mutation couldn’t have occurred 5 generations ago but it could just as well have occurred between you and your mother and in any other generation. Entirely serendipitous.
Very interesting. I learned something here. I wonder how common or rare that type of mutation is within one or two generations. Is it possible your line has an inherited tendency for such mutations, since in general MtDNA is so stable over 1,000 of years?
Excellent point, Carl. Judy, are you saying that your differences are not at one of the rapidly changing positions?
Again, the main point here is that while each of us is more likely to be somewhere near the average for any spread of results, some people have to have outcomes near the extremes, and it could very well be us. GD=0 could be there when our common ancestor is 20 generations back. And GD=2 might be just two generations back.
You may find this of interest:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4191934/
Also see this piece with a section on MtDNA: https://www.ucl.ac.uk/mace-lab/debunking/understanding
If you google the question you will find multiple articles on mutation rates.
Judy, thank you for explaining the GD0 myth. I have one individual who is 100% HVR1 and HVR2 to me. I was told that this individual may be a person who we both share a common Maternal ancestor but that we have to do a full mtDNA test to confirm. I also have several other GD1 or GD2 matches but according to other blogs/articles, they said not to put much stock in them.
Thank you for explaining and showing how your own sister and Aunt show GD1 differences.
Bruce – You definitely need to do a full sequence for MtDNA to be very useful. However, even a zero or 1 mutation may or may not help with finding a match within a genealogical time frame. It is very useful in answering specific questions, such as, was this child the child of wife one or two – unless they were descendants with a common female direct line ancestor. I used it to prove that my mother/gm/ggm were direct line descendants from the same female line – annoying my cousin who said I was just trying to be part of this family because he had never heard of my ggm.
Thank you Judy. I did my mtDNA but don’t seem to know what to do with the results. This helped explain some of it.
I have two instances in my family with no zeroes but several dozen GD1s. I considered that my family member’s single mutation difference was very recent and that they are actually part of the larger group one mutation away. In both cases, I had a fairly close cousin test and that cousin was GD0 with the larger group, proving what I had suspected. When examining the extra mutation in each of the original family members, I learned that they were both hereroplasmies.
What is heteroplasmy and how do you know your MtDNA has it?
A heteroplasmy by definition, is “the presence of a mixture of more than one type of … mitochondrial DNA … within a cell or individual.” You will see it whenever, in the results page, under Differences from (your chosen reference sequence), the code says something other than A, C, G or T.
Thank you Judy. I have also heard that it is not relevant to consider anyone that has a mutation and wanted to speak up when I heard Diahan Southard speak at a conference several years ago making this claim. I also have 1 mutation that is a heteroplasmy that appeared in a full sequence test I did along with a first cousin once removed. At first I was freaked out – that would mean my paper trail was wrong, even though I had spoken with and met family members of my mother’s generation. Fortunately, I had invited Debbie Wayne Parker to speak at OGS. She looked at my results and told me what it was which I verified on FTDNA. I also learned that some mutations of the kind I have can revert back in the next generation. It would cost a fortune to try to track down when the heteroplasmy occurred and only my sister and I survive in my direct female line from my grandmother forward – altho my brother’s DNA is stored with FTDNA. The only thing I would gain by testing the three of us would be to know if we all have the mutation.
If we did would that make it more likely that the mutation occurred with my mother or grandmother? Is there really any need to discover when it occurred genealogically speaking?
What is the mutation rate for the base locations scanned of mt DNA? There must be an average mutation rate calculated. I see scientific studies estimate ancient ages. I know averages are just averages but this would help figure out confidence probabilities.
Great question and one to which I don’t have the answer. Maybe Family Tree DNA does.
Thank you for this article ! It is one of my favorite browser bookmarks, and every time I see this very topic in a forum I promptly provide the URL to this piece.
Judy, gotta tell you again, this is one of the BEST articles I’ve read on the subject. Just got a new mtDNA match with GD=1. His research was stuck in Massachusetts and I helped him trace his maternal line back into Ireland. His earliest known maternal ancestor lived practically within shouting distance of mine. Our haplogroup is a rare one. The geographic proximity in Ireland cannot be a coincidence.